4-1 Chemical compositions
To understand the chemical states and chemical compositions of the vanadium ions at the surface sites and in the bulk lattice of the doped TiO2, the V-doped TiO2 samples calcined at different temperatures were characterized using SIMS, ICP-MS, XPS, XAS, and EPR.
Table 4-1 lists the bulk and surface V/Ti ratios of 1.00 and 0.01 at.% V-doped TiO2. The total V/Ti ratios in the 1.00 at.% V-doped TiO2 calcined at 200 and 500 °C were 1.03 × 10-2, which were similar to the added one, indicating the non-volatility of the vanadium ions in the TiO2 matrix. Although the surface V/Ti ratio at 200 °C (1.17 × 10-2) was close to the total ratio (1.03 × 10-2), the surface V/Ti ratio increases when the calcination temperature increased and reaches 7.6 times when the calcination temperature increased to 600 °C (8.86%). The total V/Ti ratios in the 0.01 at.% V-doped TiO2 calcined at 200 and 500 °C were 1.15 × 10-4 and 1.04 × 10-4 , respectively, also indicating the non-volatility of the vanadium ions in the TiO2 matrix. The surface V/Ti ratios in the 0.01 at.% V-doped TiO2
increases 2.95 times when the calcination temperature increased from 200 (1.15 × 10-4) to 600 °C (3.39×10-4), which were similar to the 1.00 at.% V-doped TiO2. This phenomenon reveals the diffusion of the vanadium ions from the inside lattice to the surface. Davidoson and Che[29] reported that metal ions migrate to the surface of matrix above the Tammann temperature at which the thermal vibrations of cations are strong enough for lattice diffusion.
Since the Tammann temperature of V2O5 is 209 °C, some vanadium ions migrated from bulk lattice toward to surface lattice of TiO2 above this temperature in this study.
Figure 4-1 shows the V 2p XP spectra of the 1.00 at.% and 4.00 at% V-doped TiO2
calcined at different temperatures. The V 2p photoelectron lines of the 1.00 at.% V-doped TiO2 were insignificant after calcination at 200 °C. Whereas, the sample showed the V 2p3/2
31
and 2p1/2 peaks centered at 517.5 and 524.3 eV, respectively, at 700 °C.[54] Similar result was found in the 4.00 at.% V-doped TiO2 sample at 200 °C. These phenomena reveal that V5+ ions were mainly at the TiO2 surface, and again prove that their concentration increased with calcination temperature. Figure 4-2 shows the V K-edge XA spectra of the 1.00 at.%
V-doped TiO2. The XAS shows the pre-edge absorptions of the V3+, V4+ and V5+ ions at 5469.9, 5470.3 and 5470.6 eV, respectively, indicating the reduced V ions within the TiO2
lattice.[35, 55-57] The intensity variation is noteworthy. The intensity increases with increasing calcination temperature. The result shows the square-pyramidal symmetry (as in V2O5) gradually formatted when increasing calcination temperature.[58] The rising-edge energy values shift from 5481.5 to 5482 eV. The result indicates the valence state of vanadium transformed to higher state.[56, 59-61] The V3+ and V4+ ions contributed a large portion at low temperatures, and they were gradually transformed to the V5+ state at elevated temperatures. Since the transformation occurred along with the thermal induced migration of the V ions, the oxidation of V3+/V4+ was possibly resulted from their interaction with O2 in the atmosphere when the reduced ions diffused to the surface. The XAS shows the near-edge appeared two peaks at 5492 and 5550 eV when calcination temperature above 600
°C. According to the literature, at temperature above to 600 °C, the V4+ ions are incorporated within the rutile structure.[62]
Figure 4-3 shows the EPR spectra of the 1.00 at.% V-doped TiO2 at different calcinations temperatures in the dark at 77K. In this study, the hyperfine structure of V4+ ions were resulted from the interaction between the vanadium nucleus spin (I=7/2) and the unpaired 3d1 electron. The peaks at 300 and 400 °C show the following EPR parameters: g⊥=1.959, A⊥
=57 G and g∥=1.917 , A∥=173 G. [63-65] This results indicate that there are V4+ ions in the anatase phase of the V-loaded sample. When the calcination temperature increases from 300 to 600 °C, the intensity of the peaks are decreased. The results indicate that the phase transformed to the rutile phase. However, the peaks at 500 and 600 °C show another
32
following EPR parameters: gxx ,gyy =1.913, Axx = 31 G, Ayy= 43 G and gzz =1.956 , Azz =152 G.[63, 66] This results indicate that there are V4+ ions in the rutile phase of the V-loaded sample. Because the XPS spectrum indicates the vanadium on the surface are V5+ ions, V4+
ions are indicated in the bulk of TiO2.[62] Figure 4-4 shows the EPR spectra of the 0.01 at.% V-doped TiO2 at different calcinations temperatures before and after UV irradiation at 77K. The peaks at 300 and 400 °C show the following EPR parameters: g⊥=1.986.[67]
This results indicate that there are V4+ ions in the interstitial sites in the anatase phase of the V-loaded sample. This signals decrease when the calcination temperature increased from 300 to 600 °C. The results indicate that the phase transformed to the rutile phase.
However, The peaks at 500 and 600 °C indicate another following EPR parameters: gxx ,gyy
=1.913, Axx = 31 G, Ayy= 43 G and gzz =1.956 , Azz =152 G. This results show that there are V4+ ions in the rutile phase of the V-loaded sample. Table 4-2 lists the g=factor of EPR.
33
Table 4-1 The bulk and surface V/Ti atomic ratios of doped TiO2.
Added V/Ti ratios calcination temperature Bulk V/Ti ratios (ICP-MS)
Surface V/Ti ratios (SIMS)
1.00 at.% V/Ti 200 °C 1.03×10-2 1.17×10-2
1.00 at.% V/Ti 300 °C 1.08×10-2 2.08×10-2
1.00 at.% V/Ti 400 °C 1.07×10-2 4.48×10-2
1.00 at.% V/Ti 500 °C 1.03×10-2 8.51×10-2
1.00 at.% V/Ti 600 °C 1.14×10-2 8.86×10-2
0.01 at.% V/Ti 200 °C 1.15×10-4 1.15×10-4
0.01 at.% V/Ti 300 °C 0.94×10-4 1.21×10-4
0.01 at.% V/Ti 400 °C 1.38×10-4 1.89×10-4
0.01 at.% V/Ti 500 °C 1.04×10-4 2.56×10-4
0.01 at.% V/Ti 600 °C 1.26×10-4 3.39×10-4
34
526 524 522 520 518 516 514
6000
5450 5460 5470 5480 5490 5500 5510 0.0
35
2800 3000 3200 3400 3600 3800 4000 4200 0
Figure 4-3 EPR spectra of 1.00 at.% V-doped TiO2 at different calcination temperature at 77K in the dark.
Figure 4-4 EPR spectra of 0.01 at.% V-doped TiO2 at different calcination temperature at 77K in the dark.
gzz =1.956, Azz =152 G gxx ,gyy =1.913 Axx = 31 G, Ayy= 43 G
2800 3200 3600 4000 4400
400 oC
36
Table 4-2 EPR Parameters of Paramametic Swcies in the pure TiO2 and V-doped TiO2.
g factor assignment Ref.
EPR Parameters of Ti3+ (electron center) Radicals
g1=1.990; g2=1.990; g3=1.960 Ti3+ (hydrated anatase) [68]
g1=1.961; g2=1.992; g3=1.992 Ti3+ (colloidal TiO2) [69]
g⊥=1.990; g∥=1.957 Ti3+ (anatase) [70]
g⊥=1.975; g∥=1.940 Ti3+ (rutile) [70]
g⊥=1.925; g∥=1.885 Ti3+ ( surface Ti3+ in colloidal TiO2 ) [71]
EPR Parameters of Oxygen Related Signals (hole center) g1=2.004; g2=2.009; g3=2.023
g⊥=2.022; g∥=2.004
g1=2.003; g2=2.008; g3=2.035
organic peroxyl (e.g., ROO．) carboxyl radical of cysteine (CH3)3N+CH2OO．
37
EPR Parameters of V4+ sites in TiO2
g⊥=1.981; g∥=1.924
g1=1.906; g2=1.960; g3=1.960 g1=1.917; g2=1.959; g3=1.959
V4+ (in anatase) [63, 67]
g1=1.913; g2=1.913; g3=1.956 g1=1.912; g2=1.914; g3=1.956
V4+ (in Rutile) [63, 66]
g1=1.923; g2=1.986; g3=1.965 g1=1.923; g2=1.967; g3=1.940
V4+ (in Interstitial Sites in Anatase) [63]
38
4-2 Microstructures
To examine the effect of the thermal-induced migration on the crystalline phase, grain size, specific surface area and bandgap, the samples were analyzed by using XRD and BET.
Figure 4-5 to Figure 4-7 shows the XRD patterns of the pure, 1.00 at.% and 0.01 at.%
V-doped TiO2 at different calcination temperatures, respectively. The anatase and rutile phase were indentified from the their typical (101) and (110) diffraction peaks at 25.4 and 27.5° 2θ positions, respectively. Table 4-3 lists the crystalline structures, surface areas and bandgaps of the pure and V-doped TiO2 after the calcination at different temperatures. The pure TiO2 contained anatase phase above 200 °C. The phase transformation from anatase to rutile took place at 600 °C. However, the incorporation of V ions into the TiO2 lattice accelerated the phase transformation and resulted in a lower phase transit temperature of 500
°C. The decreased stability of anatase phase was presumably due to the formation of V2O5
on the surface. The differential weight loss curves (shown in Figure 4-8) of the V-doped TiO2 show two peaks at 606 and 830 °C for 1.00 at.% and 0.01 at.% V-doped TiO2, respectively, indicating the formation of V2O5 moiety.[74-76] We further used GIXRD to analyze the surface structures of the doped TiO2 and the results were shown in Figure 4-9 and Figure 4-10. The 1.00 at.% V-doped TiO2 contained a weak V2O5 diffraction at 19.12° 2θ position at 300 °C. This V2O5 diffraction peak became intensive at 600 °C, which is in agreement with the observation in the thermogravimetric analysis data. Amores and Balikdjian provided a model of “sintering-induced phase transition” to describe the thermal behavior that V ions lowered phase transit temperature.[77-79] In their model, surface V species causes inefficient heat dispersion during sintering and promotes nucleation of rutile phase at the surface boundaries. Thermal induced coalescence causes the surface areas of the pure TiO2 decreasing from 131 to 1 m2/g as the temperature increased form 200 to 600 °C.
The V-doped TiO2 exhibited similar surface areas (1-124 m2/g for 0.01 at.% V-doped TiO2;
39
2-135 m2/g for 1.00 at.% V-doped TiO2) at the same temperature range. Incorporation of V ions only slightly inhibited the sintering effect because of their low concentrations. The 0.01 at.% V-doped TiO2 didn’t exist any V2O5 diffraction at 19.12° 2θ position. Low concentrations of V ions don’t have enough capacity to lead to the formation V2O5 moiety does not have enough capacity from 200 to 600 °C.
Figure 4-11 schematically illustrates the microstructures of the V-doped TiO2
transformed from low to high calcination temperature. At lower temperatures, V3+/V4+ ions disperses homogenously and interstitially within the TiO2 lattice, while the V5+ ions mainly stay at the surface. Thermal treatment induces the migration of V3+/V4+ ions moving from the inside TiO2 matrix to its surface lattice and transforming to V5+ ions. As the concentration of the accumulated V5+ ions over its solubility in the TiO2 matrix, they segregate from the TiO2 matrix to form V2O5 moiety.
Figure 4-5 The XRD patterns of the pure TiO2 at different calcination temperatures.
40
Figure 4-6 The XRD patterns of the 0.01 at.% V-doped TiO2 at different calcination temperatures.
Figure 4-7 The XRD patterns of the 1.00 at.% V-doped TiO2 at different calcination temperatures.
41
Table 4-3 The crystalline properties, surface areas and bandgaps of the pure TiO2 and V-doped TiO2 calcined at different temperatures.
Samples calcination
42
200 400 600 800
830 oC 606 oC 1.00 %V/Ti
0.01 %V/Ti
TiO2
Differential weight loss (A. U)
Temperature (oC)
Figure 4-8 The differential weight loss curves of the pure, the 0.01 at.% and 1.00 at.%
V-doped TiO2.
20 30 40 50 60 70 80
0 200 400 600 800
1000 V2O5 (19.12o)
600 oC 500 oC
400 oC 300 oC 200 oC
Intensity (A.U.)
2θ
Figure 4-9 The GI-XRD patterns of the 1.00 at.% V-doped TiO2 at different calcination temperatures.
43
20 30 40 50 60 70 80
0 200 400 600 800 1000 1200 1400
600 oC 500 oC
400 oC 300 oC 200 oC
Intensity (A.U.)
2θ
Figure 4-10 The GI-XRD patterns of the 0.01 at.% V-doped TiO2 at different calcination temperatures.
Figure 4-11 The microstructures of the V-doped TiO2 transformed from low to high calcination temperatures.
Calcination temperature V3+ and V4+ ions which were
dispersed in the TiO2 lattice V2O5
44
4-3 UV-Visible absorption
To examine the electronic structures of the photocatalysts, the optical properties of the pure and V-doped TiO2 samples were characterized in terms of UV-vis diffuse reflectance spectroscopy (DRS). Figure 4-12 displays the optical absorbance of the pure TiO2 and the 0.01 at.% V-doped TiO2 at different calcination temperatures from wavelength of 900 to 200 nm. The bandgap energy of TiO2 ranged 3.1-3.3 eV (absorption edge at 376-400 nm) below 600 °C, corresponding to the anatase.[17] There were two bands below 405 nm. One broad band was ranged between 300-350 nm and centered at 340 nm which was denoted to lower CB.[80, 81] The other band centered at 233 nm was indicated to upper CB. When the calcination temperature was higher than 600 °C, the bandgap energy of TiO2 shifted to 3.0 eV (413 nm) and lower CB center was ranged between 300-420 nm shifted to 362 nm because of generation of rutile phase decreased the bandgap range.[82] The result can be ascribed to the formation of larger particles size therefore it decrease the quantum size effect.[82, 83]
The spectra of the 0.01 at.% V-doped TiO2 shows similar absorption behavior relative to the pure TiO2. The effect of the V ions with trance amounts on the electronic structure of TiO2
sample was little to be detected. Figure 4-13 displays the optical absorbance of the 1.00 at.% V-doped TiO2 at different calcination temperatures. The bandgap energy of the 1.00 at.% V-doped TiO2 extended to 1.6 eV (779 nm). Chang et al.[35] reported that the occupied states of the V3+ and V4+ ions are located at 0.43 eV and 1.0 eV, respectively, below the bottom of the conduction band. Thus, the long wavelength absorption could be resulted from the conduction band →semi-occupied V4+ state transition or the band-tail transition.
This phenomenon of narrow bandgap to band-tail transitions result from the presence of amorphous structures in the V-doped TiO2 grain-boundaries.[35] Although some V2O5
crystals were distributed at the TiO2 surface, their contents were too small to be detected by UV-vis spectrum. Chang et al.[35] showed that the occupied states of the V3+ and V4+ ions
45
are located at 0.43 eV and 1.0 eV, respectively, below the bottom of the conduction band.
The wavelengths for the V3+/V4+ to conduction band are 1240 to 3100 nm. Therefore, the wavelengths were too long to be measured in this study. Figure 4-14 illustrates the possible electronic structure for the 1.00 at.%V-doped TiO2. Chang at al.[35] indicated that the V3+
and V4+ ions in the TiO2 crystals lead to electron-hole recombination.
46
(a)
200 300 400 500 600 700 800 900
0
Kubelka-Munk (A.U.) Kubelka-Munk (A.U.)
47
200 300 400 500 600 700 800 900
5 10 15 20
700oC 600oC
500oC
400oC 300oC 200oC
Intensity
Wavelength(nm)
Figure 4-13 UV-Vis DRS spectra of the 1.00 at.% V-doped TiO2.
Figure 4-14 The electronic structure of 1.00 at.% V-doped TiO2. V3+
+ + + +
- - -
-V4+
VB
CB 1 eV 0.43 eV
hν
Kubelka-Munk (A.U.)
48
4-4 Photocatalytic activity
The photocatalytic activities of the pure TiO2 and V-doped TiO2 were examined in terms of the degradation of 0.01 mM RhB. Figure 4-15 shows the rate constants of the pure and the V-doped TiO2 samples calcined at different temperatures. At 300-600 °C, the activity of the TiO2 samples increased from 0.0216 to 0.445 with increasing temperatures. Similar trend in the activities was found in the V-doped TiO2-based systems till 500 °C. The 0.01 at.% and 1.00 at.% V-doped TiO2 showed their highest photocatalytic activities of 0.24 and 0.019 1/min, respectively, at 500 °C. The activities of both the doped TiO2 turned down at 600 °C. The temperature dependent activity was mainly governed by crystallinity and phase compositions. Smaller amounts of defects resulted at higher crystallinity led TiO2
performing better photocatalytic efficiency. In addition, anatase and rutile composites promoted the activity because of well charge separations.[17, 20] At 300-500 °C, 0.01 at.%
V-doped TiO2 exhibited higher activity than the pure TiO2, indicating that trace amounts of V3+/V4+ ions within the lattice preserve larger numbers of effective charge carriers for photocatalysis. In contrast, the 1.00 at.% V-doped TiO2 performed the lowest activity. The lattice V3+/V4+ ions either trap electrons or holes. The trapped charge carries can not escape from the trapping sites and eventually annihilate inside the TiO2 matrix. Trace amounts of the impurities consume few charge carries from the bands to allow the remaining carriers successfully diffusing to the surface for interfacial transfer. However, over amounts of defects lead to severe electron-hole recombination, thus remarkably inhibiting the photocatalytic efficiency.[35]
For examining the photoreduction activities of the V-doped TiO2 the samples calcined at 400-600 °C were selected because their efficiencies for the photo-oxidation of RhB were higher than the samples calcined at the other temperatures. All these reactions were carried out under the same irradiation condition as that for photocatalytic oxidation of RhB. CH4
49
was the only product which was detectable in this study. Figure 4-16 shows the accumulated amounts of CH4 in the presence of the pure and the 0.01 V-doped TiO2. The pure TiO2 calcined at 500 and 600 °C exhibited similar trend in the CH4 generation. The yield of CH4 rose fast in the first hour and reached almost a steady state till the eight hour.
Calcination at 500 °C resulted the pure TiO2 in a higher reductive activity than calcination at 600 °C. The CH4 yield of the TiO2 at 500 and 600 °C at the first hour was 0.66 and 0.54 μmol/g, respectively. The TiO2 calcined at 400 °C performed a relatively low CH4 yield (0.39 μmol/g) at the first hour. However, its CH4 yield continuously increased to 0.66 μmol/g after 2 hours. After that, the yield decreased with the irradiation time and reached 0.58 μmol/g after 8 hr irraditation. Because substantial CO2 and H2O vapor still existed in the photoreductive system, the inhibited generation of CH4 reveals its fast re-oxidation.
Compared to the TiO2 calcined at 400 °C, the sample calcined at higher temperatures (500-600 °C) showed higher initial activities for CO2 reduction and retarded re-oxidation.
The photoreductive behavior of the 0.01 at.% V-doped TiO2 samples calcained at 500
°C was similar to that of the pure TiO2 at the same temperature. It produced 0.54 μmol/g of CH4 yield in the first hour and kept a steady accumulated yield in the last 7 hours. The sample calcined at 600 °C resulted in 0.35 μmol/g of CH4 yield in the first hours. The accumulated yield constantly increased with the irradiation time and reached 0.52 μmol/g at the eighth hour. Calcination at 400 °C exhibited the highest accumulated yield (0.57 μmol/g) at the third hour. Afterwards, the yield was maintained till the eight hour.
The yield of CH4 in the pure TiO2 and 0.01 at.% V-doped TiO2 systems either kept similar or decreased after the maximum has reached. However, the 1.00 at.% V-doped TiO2
samples show different phenomenon. The samples calcined at 400-600 °C all continuously increased the CH4 yield with the irradiation time (Figure 4-17). The yields of CH4 reached 0.77, 1.17 and 0.94 μmol/g in the presence of the samples calcined at 400, 500 and 600 °C, respectively.
50
Quantum efficiency, which is the ratio between the moles of products and moles of photons incident, is generally used to universally evaluate the photocatalytic performance of a certain photocatalyst and system design. Since eight moles of electrons are required to produce 1 mol of methane from CO2, the quantum efficiency of a photocatalyst for photocatalytic conversion of CO2 to CH4 is expressed as Equation 4-1.
The QE in the first hour is taken to compare the initial activities of the photocatalysts, while the QE after 8-hr irradiation was referenced to understand the reductive behavior. Table 4-4 lists the quantum efficiency of the pure TiO2 and V-doped TiO2. Figure 4-18 shows the quantum efficiency of the pure TiO2, 1.00 and 0.01 at.% V-doped TiO2 after 1-hr and 8-hr irradiation. The pure TiO2 photocatalyst calcined at 500 °C exhibited the highest QE of 2.98
% over the photocatalysts prepared under different conditions in this study. Table 4-5 lists some references of photoreduction of CO2. Li et al.[84] synthesized mesoporous silica supported Cu/TiO2 nanocomposites and carried out photoreduction experiments in a continuous-flow reactor using water vapor and CO2 under UV-light irradiation. The high surface area mesoporous silica substrate enhanced CO2 photoreduction and the QE of CH4
reached 0.28%. Wu et al.[6] designed catalyst-coated fibers to transmit and spread light inside the reactor under UV irradiation. The rate of the yield of methanol was 4.12 μmole/g-cat h and the QE reached 0.00013%. Varghese et al.[85] used N-doped titania nanotube arrays to converse CO2 and water vapor to hydrocarbons by outdoor global AM 1.5 sunlight. This hydrocarbon production rate was 111 ppm cm-2 h-1 and QE reached 0.74%.
The highest QE in this study is significantly high compared to the references. Incorporated V ions into the TiO2 lattice either at 0.01 at.% or 1.00 at.% inhibited the activities. The 0.01 at.% V-doped TiO2 inhibited the initial activities of the pure TiO2 samples calcined at 400, quantum efficiency (%) = × 100 (4-1) 8 × moles of methane yield
moles of UV photon absorbed by catalyst
51
500 and 600 °C by 0.97, 0.82 and 0.65 fold, respectively, while the 1.00 at.% V-doped TiO2
inhibited the activities by 0.92, 0.89 and 0.74 fold, respectively. However, the quantum efficiency in V-doped TiO2 showed higher QEs after irradiation of 8 hours. Relative to the pure TiO2 calcined at 400, 500 and 600 °C, the 1.00 at.% V-doped TiO2 improved the QE of CH4 by 1.33, 1.68 and 1.91 fold, respectively. These reduction results reveal that the activity of the photocatalysts evaluated in terms of their oxidation can not be referenced to predict their activity for reduction. The pure TiO2 calcined at 600 °C, which comprised of anatase and rutile phase, showed the highest oxidation efficiency for RhB, but exhibited lower activity than the sample calined at 500 °C for CO2 reduction. Moreover, heavy doping of V ions in the TiO2 lattice greatly reduced the oxidative activity. In contrast, the reductive activity of 1.00 at.% V-doped TiO2 is similar to that of doped TiO2 with 0.01 at.%
V-ion loading. These phenomena imply that the surface properties determine the photoreductive kinetics of the catalysts in stead of bulk microstructures. In addition, formation of V2O5 moieties at high-temperatures prevents the reoxidation of CH4. To further explore the surface reactions, the species generated on the samples after irradiation with UV light under different atmospheres was characterized using EPR.
It is found that the color of the photocatalysts changed during the photoreductive reactions. Figure 4-19 and 4-20 show the photographs of the photocatalysts before and after the reactions. All the photocatalysts turned to grey after the reactions. Interestingly, the grey color became lighter quickly when the photocatalysts stopped photocatalysis and were
It is found that the color of the photocatalysts changed during the photoreductive reactions. Figure 4-19 and 4-20 show the photographs of the photocatalysts before and after the reactions. All the photocatalysts turned to grey after the reactions. Interestingly, the grey color became lighter quickly when the photocatalysts stopped photocatalysis and were